Is it legal to only ask your happy clients for a review?

What did Meriton actually do? #

In 2018 the Federal Court fined Meriton Apartments $3 million. Not for writing fake reviews. Not for deleting real ones.

Meriton sent guests a follow-up email that triggered a TripAdvisor review prompt. Where staff knew something had gone wrong during a stay, a broken lift, building noise, no hot water, they altered the guest's email address before it went out. The message never arrived, TripAdvisor never sent the prompt, and the unhappy guest never got the chance to review.

Everything Meriton knew about those guests, they knew legitimately. Every review that did get published was real. The only thing they did was decide who got asked.

That cost them three million dollars.

Is it different if a person decides rather than the software? #

No, and this is the part that catches good firms out.

Meriton had no feature called "review gating". It had staff making a judgement call, one guest at a time, using information they were perfectly entitled to have. Which is exactly the two-step version everybody thinks is the clever workaround: run a satisfaction survey, read the results, then send review requests to the people who scored well.

Doing it by hand does not launder it. If anything it is worse, because now there is a record. A satisfaction dashboard showing who was unhappy, sitting next to a request log showing who got asked, is not an internal tool any more.

So what am I actually allowed to do? #

Here is the line, and it is simpler than most people expect.

You can choose when someone gets asked. You cannot choose whether, based on what you think they will say.

That means you can ask at the end of a matter rather than the middle. You can ask after the settlement, the lodgement, the handover. You can decide every client enters the review cycle thirty days after completion. All fine. What you cannot do is look at a score and quietly move someone off the list.

Will asking everyone drag my rating down? #

A little. And that is fine, because the average is the wrong number to protect.

Northwestern's Spiegel Research Center looked at how ratings affect purchase likelihood and found it peaks somewhere around 4.2 to 4.5 stars, then falls as the score climbs toward 5.0. Across every category they examined, there was not a single one where the optimal rating was a perfect five. A flawless score reads less like excellence and more like a small, tidy conspiracy.

What actually moves the needle is volume and recency. Forty reviews averaging 4.6, most of them from this year, beat eight immaculate fives from 2023 on every measure that matters: search ranking, AI citation, and whether a stranger believes you.

Gating protects the average and starves the volume. It is the worst trade in the category.

What this looks like on a Tuesday #

Three moves, none of which require you to filter anyone.

Ask everyone, at the moment of maximum goodwill

How someone feels when you ask is mostly a function of when you ask. Right after the win, the lodgement, the problem solved. Same request, different Tuesday, much better answer. That is timing, and timing is allowed.

Recover, then ask

An unhappy client tells you something is wrong. You ring them that afternoon and fix it. They go into the next cycle with everybody else. Recovered clients are often your best reviewers, because a story about a problem you sorted is more persuasive than a story about nothing going wrong.

Make it a system, not a decision

The real reason firms gate is that asking is awkward, so they would rather only do the comfortable ones. Put the link in the invoice template and the email signature and the ask stops being a judgement call. Nobody has to decide anything, which is the entire point.

Why I will not build the filter #

Every client who sees our review system asks for the same feature. Can it only send to the happy ones. It is the most requested thing in the product, and it is the one thing I have deliberately taken out. There was a gated mode in the schema. I deleted it.

Partly that is because I am not going to hand an accounting or law firm a practice that breaches the ACL and their own advertising rules in the same afternoon.

But mostly it is this. Gating is how a firm that is not the best buys the look of being the best. If everyone filters, a five-star average stops telling you who does good work and starts telling you who is diligent about screening. The firm that gets punished is the one quietly doing excellent work and asking everybody, honestly.

I build systems so the best firm wins, not the best connected one. That is the whole point, and it does not survive a filter.

So the review system I build for clients asks everyone, or it asks nobody and keeps the feedback internal for service recovery. There is no third setting, and there will not be one.

None of this is legal advice, and if you are in a regulated profession it is worth ten minutes with someone who is. But the $3 million is real, and it was spent on a shortcut.